3. FEEDFORWARD, NOT FEEDBACK: instead of having students submit a paper and return a grade (with maybe a little bit of feedback), I now have students submit a first complete draft, for which I give them a provisional grade and feedforward that they can use to revise and resubmit for a final grade.
May 7, 2025 at 1:41 PM
3. FEEDFORWARD, NOT FEEDBACK: instead of having students submit a paper and return a grade (with maybe a little bit of feedback), I now have students submit a first complete draft, for which I give them a provisional grade and feedforward that they can use to revise and resubmit for a final grade.
Okay, but with *enough* Ti-84s - github.com/dmhacker/cyb...
GitHub - dmhacker/cyberdyne: Feedforward neural network in a TI-84 graphing calculator
Feedforward neural network in a TI-84 graphing calculator - dmhacker/cyberdyne
github.com
September 22, 2026 at 2:38 AM
Okay, but with *enough* Ti-84s - github.com/dmhacker/cyb...
"...feedforward proviral circuit..."
kinda encapsulates the public health response, too!
kinda encapsulates the public health response, too!
September 20, 2026 at 9:11 AM
"...feedforward proviral circuit..."
kinda encapsulates the public health response, too!
kinda encapsulates the public health response, too!
A "feedforward proviral circuit" will emerge in a school soon
September 20, 2026 at 3:12 PM
A "feedforward proviral circuit" will emerge in a school soon
Excited about our new preprint: “The illusory simplicity of the feedforward pass: evidence for the dynamical nature of stimulus encoding along the primate ventral stream”
arxiv.org/abs/2604.12825
Work with Sushrut Thorat, Anna Mitola, Paolo Papale, Peter König & Tim Kietzmann
🧵 thread below
arxiv.org/abs/2604.12825
Work with Sushrut Thorat, Anna Mitola, Paolo Papale, Peter König & Tim Kietzmann
🧵 thread below
The illusory simplicity of the feedforward pass: evidence for the dynamical nature of stimulus encoding along the primate ventral stream
In studying primate vision, a large body of work focuses on the first feedforward sweep. During this initial time window, information is thought to pass through ventral stream regions in a stage-like ...
arxiv.org
May 11, 2026 at 10:50 AM
Excited about our new preprint: “The illusory simplicity of the feedforward pass: evidence for the dynamical nature of stimulus encoding along the primate ventral stream”
arxiv.org/abs/2604.12825
Work with Sushrut Thorat, Anna Mitola, Paolo Papale, Peter König & Tim Kietzmann
🧵 thread below
arxiv.org/abs/2604.12825
Work with Sushrut Thorat, Anna Mitola, Paolo Papale, Peter König & Tim Kietzmann
🧵 thread below
the only use for the word 'feedforward' is when u buy an extra order of fries for the person behind u in the drive thru
September 14, 2023 at 3:31 PM
the only use for the word 'feedforward' is when u buy an extra order of fries for the person behind u in the drive thru
Categorization is not just feedforward and at the "end" of cortex. It is an interplay between feedforward and feedback signals and areas.
Dynamic reversal of IT-PFC information flow orchestrates visual categorization under perceptual uncertainty
doi.org/10.1016/j.pn...
#neuroscience
Dynamic reversal of IT-PFC information flow orchestrates visual categorization under perceptual uncertainty
doi.org/10.1016/j.pn...
#neuroscience
Redirecting
doi.org
August 5, 2026 at 3:50 PM
Categorization is not just feedforward and at the "end" of cortex. It is an interplay between feedforward and feedback signals and areas.
Dynamic reversal of IT-PFC information flow orchestrates visual categorization under perceptual uncertainty
doi.org/10.1016/j.pn...
#neuroscience
Dynamic reversal of IT-PFC information flow orchestrates visual categorization under perceptual uncertainty
doi.org/10.1016/j.pn...
#neuroscience
A huge effort and a new take on what constitutes the "feedforward pass" across the visual system. Make sure to check out our new preprint:
Excited about our new preprint: “The illusory simplicity of the feedforward pass: evidence for the dynamical nature of stimulus encoding along the primate ventral stream”
arxiv.org/abs/2604.12825
Work with Sushrut Thorat, Anna Mitola, Paolo Papale, Peter König & Tim Kietzmann
🧵 thread below
arxiv.org/abs/2604.12825
Work with Sushrut Thorat, Anna Mitola, Paolo Papale, Peter König & Tim Kietzmann
🧵 thread below
The illusory simplicity of the feedforward pass: evidence for the dynamical nature of stimulus encoding along the primate ventral stream
In studying primate vision, a large body of work focuses on the first feedforward sweep. During this initial time window, information is thought to pass through ventral stream regions in a stage-like ...
arxiv.org
May 11, 2026 at 10:52 AM
A huge effort and a new take on what constitutes the "feedforward pass" across the visual system. Make sure to check out our new preprint:
Feedforward sweep is a bad idea. And brains don't seem to do it.
Excited about our new preprint: “The illusory simplicity of the feedforward pass: evidence for the dynamical nature of stimulus encoding along the primate ventral stream”
arxiv.org/abs/2604.12825
Work with Sushrut Thorat, Anna Mitola, Paolo Papale, Peter König & Tim Kietzmann
🧵 thread below
arxiv.org/abs/2604.12825
Work with Sushrut Thorat, Anna Mitola, Paolo Papale, Peter König & Tim Kietzmann
🧵 thread below
The illusory simplicity of the feedforward pass: evidence for the dynamical nature of stimulus encoding along the primate ventral stream
In studying primate vision, a large body of work focuses on the first feedforward sweep. During this initial time window, information is thought to pass through ventral stream regions in a stage-like ...
arxiv.org
May 14, 2026 at 2:10 PM
Feedforward sweep is a bad idea. And brains don't seem to do it.
This paper shows that some human "gestalt" style visual processing can emerge in purely feedforward deep nets with the right inductive biases and fine-tuning:
journals.plos.org/ploscompbiol...
🧠📈 🧪 #NeuroAI
journals.plos.org/ploscompbiol...
🧠📈 🧪 #NeuroAI
A feedforward mechanism for human-like contour integration
Author summary A central challenge in vision science is understanding how the visual system links fragmented local features into coherent object representations. One foundational process supporting th...
journals.plos.org
August 26, 2025 at 5:42 PM
This paper shows that some human "gestalt" style visual processing can emerge in purely feedforward deep nets with the right inductive biases and fine-tuning:
journals.plos.org/ploscompbiol...
🧠📈 🧪 #NeuroAI
journals.plos.org/ploscompbiol...
🧠📈 🧪 #NeuroAI
The control of movement gradually transitions from feedback control to feedforward adaptation throughout childhood
www.nature.com/articles/s41...
#sensorimotor
www.nature.com/articles/s41...
#sensorimotor
The control of movement gradually transitions from feedback control to feedforward adaptation throughout childhood - npj Science of Learning
npj Science of Learning - The control of movement gradually transitions from feedback control to feedforward adaptation throughout childhood
www.nature.com
March 14, 2025 at 12:32 PM
The control of movement gradually transitions from feedback control to feedforward adaptation throughout childhood
www.nature.com/articles/s41...
#sensorimotor
www.nature.com/articles/s41...
#sensorimotor
New theory: Categorization is Baked Into the Brain
Categorization is not a late stage of sensory processing but a core computation that begins at the earliest stages, occurring throughout the cortex as predictive feedback shapes feedforward processing.
doi.org/10.31234/osf...
#neuroscience
Categorization is not a late stage of sensory processing but a core computation that begins at the earliest stages, occurring throughout the cortex as predictive feedback shapes feedforward processing.
doi.org/10.31234/osf...
#neuroscience
OSF
doi.org
August 15, 2025 at 11:40 AM
New theory: Categorization is Baked Into the Brain
Categorization is not a late stage of sensory processing but a core computation that begins at the earliest stages, occurring throughout the cortex as predictive feedback shapes feedforward processing.
doi.org/10.31234/osf...
#neuroscience
Categorization is not a late stage of sensory processing but a core computation that begins at the earliest stages, occurring throughout the cortex as predictive feedback shapes feedforward processing.
doi.org/10.31234/osf...
#neuroscience
Introducing VGGT (CVPR'25), a feedforward Transformer that directly infers all key 3D attributes from one, a few, or hundreds of images, in seconds!
Project Page: vgg-t.github.io
Code & Weights: github.com/facebookrese...
Project Page: vgg-t.github.io
Code & Weights: github.com/facebookrese...
March 17, 2025 at 2:08 AM
Introducing VGGT (CVPR'25), a feedforward Transformer that directly infers all key 3D attributes from one, a few, or hundreds of images, in seconds!
Project Page: vgg-t.github.io
Code & Weights: github.com/facebookrese...
Project Page: vgg-t.github.io
Code & Weights: github.com/facebookrese...
Just out! Expectations modulate stimulus-evoked recurrent, but not feedforward, processing, and this is attention-dependent. NMDA-dependent feedback specifically supports perceptual integration (illusory contours), rather than mediating expectations (base rate). www.jneurosci.org/content/46/1...
Effects of Expectation, Attention, and NMDA Receptor Blockade on Feedforward and Feedback Processing
Perception is increasingly viewed as an inferential process wherein sensory inputs are integrated with prior expectations. We employed time-resolved decoding on electroencephalography (EEG) data ( n ...
www.jneurosci.org
April 7, 2026 at 2:54 PM
Just out! Expectations modulate stimulus-evoked recurrent, but not feedforward, processing, and this is attention-dependent. NMDA-dependent feedback specifically supports perceptual integration (illusory contours), rather than mediating expectations (base rate). www.jneurosci.org/content/46/1...
my feeling is this is actually a single-pass feedforward transformer arch with custom state and a classifier head clamping the output distribution. you paralellize each schema "option" into its own inference pass, return the collated batch
September 16, 2026 at 1:48 AM
my feeling is this is actually a single-pass feedforward transformer arch with custom state and a classifier head clamping the output distribution. you paralellize each schema "option" into its own inference pass, return the collated batch
Happy to share the CA3 dataset is published. First large-scale dense connectomics from hippocampus. First connectomic dataset imaged in @princetonneuro.bsky.social. A harbinger of many larger datasets to come (for both). doi.org/10.1038/s415... #hippocampus But my personal favorite is (1/2)
Hippocampal CA3 connectomics reveals a gradient of mossy fiber inputs and selective feedforward inhibition onto pyramidal cells - Nature Neuroscience
The authors present a dense connectomic reconstruction of a 0.1 mm3 hippocampal CA3 volume and reveal spatial gradients, high convergence and cell-type-selective feedforward inhibition of inputs to py...
doi.org
August 7, 2026 at 2:41 AM
Happy to share the CA3 dataset is published. First large-scale dense connectomics from hippocampus. First connectomic dataset imaged in @princetonneuro.bsky.social. A harbinger of many larger datasets to come (for both). doi.org/10.1038/s415... #hippocampus But my personal favorite is (1/2)
(link fixed) New theory: Categorization is Baked Into the Brain
It is not at the end of sensory processing but a core computation that begins at the earliest stages, occurring throughout the cortex as predictive feedback shapes feedforward processing. #neuroscience
doi.org/10.31234/osf...
It is not at the end of sensory processing but a core computation that begins at the earliest stages, occurring throughout the cortex as predictive feedback shapes feedforward processing. #neuroscience
doi.org/10.31234/osf...
OSF
doi.org
August 25, 2025 at 12:18 PM
(link fixed) New theory: Categorization is Baked Into the Brain
It is not at the end of sensory processing but a core computation that begins at the earliest stages, occurring throughout the cortex as predictive feedback shapes feedforward processing. #neuroscience
doi.org/10.31234/osf...
It is not at the end of sensory processing but a core computation that begins at the earliest stages, occurring throughout the cortex as predictive feedback shapes feedforward processing. #neuroscience
doi.org/10.31234/osf...
#Vision is often described as hierarchical and feedforward, progressing from low-level features to high-level representations. But are the early #feedforward stages truly the most essential for #perception?
Let us show you why not
jov.arvojournals.org/article.aspx... 🧵1/9
Let us show you why not
jov.arvojournals.org/article.aspx... 🧵1/9
November 17, 2025 at 2:07 PM
#Vision is often described as hierarchical and feedforward, progressing from low-level features to high-level representations. But are the early #feedforward stages truly the most essential for #perception?
Let us show you why not
jov.arvojournals.org/article.aspx... 🧵1/9
Let us show you why not
jov.arvojournals.org/article.aspx... 🧵1/9
Feedback and feedforward control are differentially delayed in cerebellar ataxia
www.biorxiv.org/content/10.1...
#sensorimotor
www.biorxiv.org/content/10.1...
#sensorimotor
February 16, 2025 at 7:50 PM
Feedback and feedforward control are differentially delayed in cerebellar ataxia
www.biorxiv.org/content/10.1...
#sensorimotor
www.biorxiv.org/content/10.1...
#sensorimotor
CO2 is an important cue used by #mosquitoes when searching for a human blood meal. @darbly.bsky.social @megyounger.bsky.social &co show that the wiring of #neurons that sense CO2 differs from that in #Drosophila, allowing robust detection of this key host cue @plosbiology.org 🧪 plos.io/4cFCPpV
September 11, 2026 at 8:50 AM
CO2 is an important cue used by #mosquitoes when searching for a human blood meal. @darbly.bsky.social @megyounger.bsky.social &co show that the wiring of #neurons that sense CO2 differs from that in #Drosophila, allowing robust detection of this key host cue @plosbiology.org 🧪 plos.io/4cFCPpV
Angiopoietin–TIE2 feedforward circuit promotes PIK3CA-driven venous malformations. Read our work in @natcardiovascres.nature.com by Marle Kraft et al.
rdcu.be/enhST
rdcu.be/enhST
Angiopoietin–TIE2 feedforward circuit promotes PIK3CA-driven venous malformations
Nature Cardiovascular Research - Kraft et al. identify TIE2 activation as a key driver of PIK3CA-related venous malformations and demonstrate that inhibiting TIE2 signaling can suppress the growth...
rdcu.be
May 25, 2025 at 2:34 PM
Angiopoietin–TIE2 feedforward circuit promotes PIK3CA-driven venous malformations. Read our work in @natcardiovascres.nature.com by Marle Kraft et al.
rdcu.be/enhST
rdcu.be/enhST
In a new Science study, researchers report on transcriptional adaptation, a dual feedback and feedforward mechanism that uses genetic redundancy to compensate for mutations in protein-coding genes.
Learn more in a new #SciencePerspective: https://scim.ag/4ryeXtN
Learn more in a new #SciencePerspective: https://scim.ag/4ryeXtN
Keeping cells fit
Fragments of aberrant cytoplasmic mRNA pair with nuclear RNAs to augment transcription
scim.ag
February 28, 2026 at 8:10 PM
In a new Science study, researchers report on transcriptional adaptation, a dual feedback and feedforward mechanism that uses genetic redundancy to compensate for mutations in protein-coding genes.
Learn more in a new #SciencePerspective: https://scim.ag/4ryeXtN
Learn more in a new #SciencePerspective: https://scim.ag/4ryeXtN
∂DSA finds lots of RNN solutions to the same integration task:
Line attractors ✅
Feedforward chains ✅
Oscillations ✅
Unstable dynamics ✅
Line attractors ✅
Feedforward chains ✅
Oscillations ✅
Unstable dynamics ✅
We also built a fast, differentiable form of DSA — ∂DSA. With this, we showed we can train RNNs online with DSA to learn novel, never-before-hypothesized solutions to computational tasks! (Stay tuned for future work on this too!) (8/)
September 21, 2026 at 8:27 PM
∂DSA finds lots of RNN solutions to the same integration task:
Line attractors ✅
Feedforward chains ✅
Oscillations ✅
Unstable dynamics ✅
Line attractors ✅
Feedforward chains ✅
Oscillations ✅
Unstable dynamics ✅
New #neuroskyence #visionscience by my excellent colleague Ecem Altan: We study higher-level processes in Ponzo-like illusions. In brief, inversion reduces the illusion & V1 activity reflects that difference, suggesting the V1 effect isn't just from feedforward processing:
doi.org/10.1098/rspb...
doi.org/10.1098/rspb...
May 14, 2025 at 1:18 AM
New #neuroskyence #visionscience by my excellent colleague Ecem Altan: We study higher-level processes in Ponzo-like illusions. In brief, inversion reduces the illusion & V1 activity reflects that difference, suggesting the V1 effect isn't just from feedforward processing:
doi.org/10.1098/rspb...
doi.org/10.1098/rspb...
By utilizing the visual backward masking paradigm, this study aimed to disentangle the contributions of feedforward and recurrent processing, revealing that recurrent processing significantly shapes the object representations across the ventral visual stream.
journals.plos.org/plosbiology/...
journals.plos.org/plosbiology/...
Recurrence affects the geometry of visual representations across the ventral visual stream in the human brain
The specific roles of feedforward and recurrent processing in human visual object recognition remain incompletely understood. In this neuroimaging and computational modelling study the authors isolate...
journals.plos.org
October 10, 2025 at 2:36 PM
By utilizing the visual backward masking paradigm, this study aimed to disentangle the contributions of feedforward and recurrent processing, revealing that recurrent processing significantly shapes the object representations across the ventral visual stream.
journals.plos.org/plosbiology/...
journals.plos.org/plosbiology/...